Apple M1 Microarchitecture Research by Dougall Johnson

Firestorm: Overview | Base Instructions | SIMD and FP Instructions
Icestorm:  Overview | Base Instructions | SIMD and FP Instructions

FCMLE (scalar, zero, S)

Test 1: uops

Code:

  fcmle s0, s0, #0
  movi v0.16b, 1
  movi v1.16b, 2

(no loop instructions)

1000 unrolls and 1 iteration

Retires: 1.000

Issues: 1.000

Integer unit issues: 0.000

Load/store unit issues: 0.000

SIMD/FP unit issues: 1.000

retire (01)cycle (02)031e3f4e51inst issue (52)~issue fp/simd (54)~dispatch fp/simd (57)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op fp/simd (7e)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst neon or fp (9a)a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)f5f6f7f8fd
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073216221786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073216221786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073216221786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073216221786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073216221786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073216221786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073216221786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073216221786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000373216221786100020382038203820382038
100420371506116862510001000100026452102018203720371571318951000100010002037203711100110000073216221786100020382038203820382038

Test 2: Latency 1->2

Code:

  fcmle s0, s0, #0
  movi v0.16b, 1
  movi v1.16b, 2

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)03181e1f3f4e51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa6a8accfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
102042003715000071196862510100100100001001000050028475211200182003720037184213187451010020010000200100002003720037111020110099100100100001000007101161119791100001002003820038200382003820038
102042003715000061196862510100100100001001000050028475211200182003720037184213187451010020010000200100002003720037111020110099100100100001000007101161119791100001002003820038200382003820038
102042003715000061196862510100100100001001000050028475211200182003720037184213187451010020010000200100002003720037111020110099100100100001000107101161119791100001002003820038200382006220038
1020420037150000399196862510100100100001001000050028475211200182003720037184213187451010020010000200100002003720037111020110099100100100001000607101161119791100001002003820038200382003820038
1020420037150000126196862510100100100001001000050028475211200182003720037184213187451010020010000200100002003720037111020110099100100100001000007101161119791100001002003820038200382003820038
1020420037150000208196862510100100100001001000050028475211200182003720037184213187451010020010000200100002003720037111020110099100100100001000007101161119791100001002003820038200382003820038
1020420037150000166196862510100100100001001000050028475211200182003720037184213187451010020010000200100002003720037111020110099100100100001000007101161119791100001002003820038200382003820038
1020420037150000103196862510100100100001001000050028475211200182003720037184213187451010020010000200100002003720037111020110099100100100001000107101161119791100001002003820038200382003820038
1020420037150000103196862510100100100001001000050028475211200182003720037184213187451010020010000200100002003720037111020110099100100100001000007101161119791100001002003820038200382003820038
102042003715000061196862510100100100001001000050028475211200182003720037184213187451010020010000200100002003720037111020110099100100100001000007101161119791100001002003820038200382003820038

1000 unrolls and 10 iterations

Result (median cycles for code): 2.0037

retire (01)cycle (02)031e1f3f4e51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9faccfd5d6ddinst fetch restart (de)e0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
10024200371500014519686251001010100001010000502847521120018200372003718443318767100102010000201000020037200371110021109101010000103640216231978610000102003820038200382003820038
1002420037150006119686251001010100001010000502847521020018200372003718443318767100102010000201000020037200371110021109101010000100640216231978610000102003820038200382003820038
10024200371500072619686251001010100001010000502847521020018200372003718443318767100102010000201000020037200371110021109101010000100640216221978610000102003820038200382003820038
10024200371500012619686251001010100001010000502847521020018200372003718443318767100102010000201000020037200371110021109101010000100640216231978610000102003820038200382003820038
1002420037150006119686251001010100001010000502847521020018200372003718443318767100102010000201000020037200371110021109101010000100640216231978610000102003820038200382003820038
1002420037149008219686251001010100001010000502847521020018200372003718443318767100102010000201000020037200371110021109101010000100640216221978610000102003820038200382003820038
1002420037150006119686251001010100001010000502847521020018200372003718443318767100102010000201000020037200371110021109101010000100640216221978610000102003820038200382003820038
10024200371500044119686251001010100001010000502847521020018200372003718443318767100102010000201000020037200371110021109101010000100640216221978610000102003820038200382003820038
10024200371500012619686251001010100001010000502847521020018200372003718443318767100102010000201000020037200371110021109101010000100640216221978610000102003820038200382003820038
1002420037150006119686251001010100001010000502847521020018200372003718443318767100102010000201000020037200371110021109101010000100640216231978610000102003820038200382003820038

Test 3: throughput

Count: 8

Code:

  fcmle s0, s8, #0
  fcmle s1, s8, #0
  fcmle s2, s8, #0
  fcmle s3, s8, #0
  fcmle s4, s8, #0
  fcmle s5, s8, #0
  fcmle s6, s8, #0
  fcmle s7, s8, #0
  movi v8.16b, 9

(fused SUBS/B.cc loop)

100 unrolls and 100 iterations

Result (median cycles for code divided by count): 0.2505

retire (01)cycle (02)03041e3f51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa8acc5branch mispredict (cb)cdcfd5d6dde0? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
80204200571501043425801081008000810080020500640132200192003820038997769989801202008003220080032200382003811802011009910010080000100001115118116020035800001002003920039200392003920039
80204200381500122925801081008000810080020500640132200192003820038997769989801202008003220080032200382003811802011009910010080000100001115118016020035800001002003920039200392003920039
80204200381500034825801081008000810080020500640132200192003820038997769989801202008003220080032200382003811802011009910010080000100001115118016020035800001002003920039200392003920039
8020420038150002925801081008000810080020500640132200192003820038997769989801202008003220080032200382003811802011009910010080000100001115118016020035800001002003920039200392003920039
80204200381500040925801081008000810080020500640132200192003820038997769989801202008003220080032200382003811802011009910010080000100001115118016020035800001002003920039200392003920039
8020420038150005225801081008000810080020500640132200192003820038997769989801202008003220080032200382003811802011009910010080000100001115118016020035800001002003920039200392003920039
8020420038150002925801081008000810080020500640132200192003820038997769989801202008003220080032200382014111802011009910010080000100001115118016020035800001002003920039200392003920039
8020420038150002925801081008000810080020500640132200192003820038997769989801202008003220080032200382003811802011009910010080000100001115118016020035800001002003920039200392008720039
8020420038150002925801081008000810080020500640132200192003820038997769989801202008003220080032200382003811802011009910010080000100001115118016020035800001002003920039200392003920039
8020420038150002925801081008000810080020500640132200192003820038997769989801202008003220080032200382003811802011009910010080000100001115118016020035800001002003920039200392003920039

1000 unrolls and 10 iterations

Result (median cycles for code divided by count): 0.2505

retire (01)cycle (02)0308091e3f51inst issue (52)~issue int (53)~issue fp/simd (54)~dispatch int (56)~dispatch fp/simd (57)huge thing int (59)huge thing fp/simd (5b)60696d6edispatch stall (70)scheduler rewind (75)scheduler stall (76)~dispatch op (78)~map op int (7c)~map op fp/simd (7e)~map lookup int (7f)~map lookup fp/simd (81)8283pipeline redirect (84)85inst all (8c)inst branch (8d)inst branch taken (90)inst b.cc (94)inst integer (97)inst neon or fp (9a)9fa1acbranch mispredict (cb)cfd5d6dadbddinst fetch restart (de)e0ec? fp/simd (ee)gpr retires (ef)f5f6f7f8fd
8002420049150000392580010108000010800005064000012001920038200389989310011800102080000208000020038200381180021109101080000100005024116001120035080000102003920039200392003920039
8002420038150000392580010108000010800005064000012001920038200389989310011800102080000208000020038200381180021109101080000100005020116001120035080000102003920039200392003920039
8002420038150000392580010108000010800005064000012001920038200389989310011800102080000208000020038200381180021109101080000100005020116001120035080000102003920039200392003920039
8002420038150010392580010108000010800005064000012001920038200389989310011800102080000208000020038200381180021109101080000100005020116001120035080000102003920039200392003920039
8002420038150000392580010108000010800005064000012001920038200389989310011800102080000208000020038200381180021109101080000100005020116031120035080000102003920039200392003920039
8002420038150000392580010108000010800005064000012001920038200389989310011800102080000208000020038200381180021109101080000100005020116001120035080000102003920039200392003920039
8002420038150000392580010108000010800005064000012001920038200389989310011800102080000208000020038200381180021109101080000100005020116001120035080000102003920039200392003920039
8002420086150001239258001010800001080000506400001200192003820038998931001180010208000020800992003820038118002110910108000010046505020116001120035080000102003920039200922008720039
80024200991502003222580010108000010800975064000012001920038200389989310011800102080000208000020038200381180021109101080000100005020116001120035080000102003920039200392003920039
8002420038150000392580010108000010800005064000012001920038200389989310011800102080000208000020038200381180021109101080000100005020116001120035080000102003920039200392003920039